AI risk class check

Which obligations of the AI Act apply to your AI system? The check classifies a system in a few minutes: prohibited practice, high-risk, transparency obligations or minimal risk, with legal references and dates as amended by the Digital Omnibus (Regulation (EU) 2026/1744). Free, no sign-up, your answers are not stored.

Check each AI system separately. The check does not cover general-purpose AI models (Chapter V) or the exclusions from the scope, for example for research or purely military purposes (Art. 2).

Your role

1. What role does your company have for this AI system?

Provider
You develop the AI system or have it developed and place it on the market or put it into service under your own name or trademark, even only internally or free of charge (Art. 3(3)). Deployer
You use an AI system under your own authority in a professional context, for example a chat assistant or an AI feature in your software (Art. 3(4)). Both
You develop the system and also use it yourself.

Prohibited practices (Art. 5)

2. Does any of these apply to the AI system? (multiple choice)

Prohibited practices under Art. 5. If none applies, leave everything empty.

Influences people subliminally or with manipulative or deceptive techniques so that they take decisions that can cause significant harm to them or others Exploits people’s age, disability or a specific social or economic situation to change their behaviour in a harmful way Generates or manipulates realistic intimate or sexual depictions of identifiable persons without their explicit consent, or child sexual abuse material Evaluates or classifies people over a period of time based on their social behaviour or personal characteristics and thereby treats them unfavourably in unrelated contexts or disproportionately (social scoring) Predicts whether a person will commit a criminal offence based solely on profiling or personality traits Builds facial recognition databases through untargeted scraping of facial images from the internet or CCTV footage Infers emotions of people in the workplace or in education institutions (except for medical or safety reasons) Categorises people based on biometric data to deduce the sensitive characteristics listed in Art. 5(1)(g), such as political opinions, religion or sexual orientation Identifies people biometrically and remotely in real time in publicly accessible spaces for law enforcement purposes

High-risk: products (Annex I)

3. Is the AI system itself a product, or a safety component of a product, covered by EU product legislation listed in Annex I?

Yes, product under section A
e.g. toys, lifts, equipment for potentially explosive atmospheres, radio equipment, pressure equipment, cableways, personal protective equipment, gas appliances, medical devices, in vitro diagnostics Yes, product under section B
e.g. machinery, motor vehicles, agricultural and forestry vehicles, two- and three-wheel vehicles, marine equipment, rail systems, civil aviation and drones No Don’t know

If yes: must the product undergo a conformity assessment by a third party (notified body)?

YesNoDon’t know

Derogations (Art. 6(3))

Questions 5 and 6 only if you ticked something in question 4.

5. Does the AI system perform profiling of natural persons, i.e. evaluate personal aspects such as work performance, reliability or behaviour?

YesNoDon’t know

6. Does one of these conditions apply?

Then a system listed in Annex III is not considered high-risk, because it does not materially influence the outcome of decision-making (Art. 6(3)).

It only performs a narrow procedural taskIt only improves the result of a previously completed human activityIt detects decision-making patterns or deviations from them, without replacing or influencing the human assessment without proper human reviewIt only performs a preparatory task for an assessmentNone of theseDon’t know

Transparency obligations (Art. 50)

7. What else does the AI system do? (multiple choice)

Transparency obligations under Art. 50. If none applies, leave everything empty.

Interacts directly with people, for example as a chatbot or voice assistant Generates synthetic audio, image, video or text content Recognises emotions or categorises people based on biometric data Generates or manipulates image, audio or video content that appears authentic and resembles real persons, places or events (deep fakes) Generates or manipulates text published to inform the public on matters of public interest

Frequently asked questions about the AI Act

What is an AI system under the Regulation?

A machine-based system that operates with some degree of autonomy and infers from its input how to generate outputs such as predictions, content, recommendations or decisions (Art. 3(1)). Rule-based software that only executes fixed rules is generally not covered.

We only use a general chat assistant. What applies to us?

You are a deployer. Then mainly AI literacy under Art. 4 applies; transparency obligations under Art. 50(4) are added if, for example, you publish deep fakes. If you use the assistant for a purpose listed in Annex III, such as pre-selecting applications, you may be considered the provider of a high-risk AI system yourself (Art. 25(1)(c)).

What did the Digital Omnibus change?

Regulation (EU) 2026/1744 has applied since 27 July 2026. The obligations for high-risk AI systems under Annex III apply from 2 December 2027, under Annex I from 2 August 2028. Art. 4 requires measures to support AI literacy instead of a specific level. Newly prohibited from 2 December 2026 are non-consensual intimate depictions and child sexual abuse material. Machinery is now listed in Annex I section B.

Are my answers stored?

No. The evaluation happens only in memory and is neither stored nor logged.

Does the check replace a legal review?

No. It gives a first orientation based on the wording of the law. For borderline cases, in particular high-risk and the derogations under Art. 6(3), we recommend a legal review.

Try Complass with your own data

Set up your trial yourself, with a sample project per framework if you like. If you would rather see Complass first, request a demo.

Try for free Request a demo